LocoNotes FAQ (frequently asked questions)

Index For This Page

This page was last updated on June 19, 2024.

(Return to the Locomotive Notes Index Page)

EMD Frame Numbers

Can someone please tell me how and where to find frame numbers on locomotives?

Starting about 1944 EMD put order and sequence numbers on every unit. They are on the outside of the frame at the right front and left rear. Imagine getting on a unit at the right front or left rear. When your hand on the handrail is even with the frame look about a foot to the left of your hand. There will be a number stamped in the frame with a sideways "1" in place of a real dash. Something like "7075-5". You DO NOT have to get on the unit to see this. Was only using that analogy to get you positioned. (Russ Stodtz)

EMC/D frame numbers on switchers and road units like GP7 to 60 etc are found in the right front and left rear of the frame. But EMC did not always stamp the frames in the early years-most SW1's from the 1939-42 era do not seem to have frame numbers, same for NW2's. On cab units like E7's F7's the frame is on the right rear, to the right of the coupler on one of the supports. (Ken Ardinger)

EMD SD7/9/18/24's have a different sort of frame with the end platform welded on. On those units the number is on the corner of that end platform facing out. Very easy to see. (Russ Stodtz)

As far as I know, the IC didn't alter the frames on the old SD24s when they went through the rebuilt program that turned them into SD20s back in the late 1970s/early 1980s. I've seen quite a few of them that have been sold off in the past few years that are now in service on the IAIS, I&M, IHB, WSOR and the NREX leasers that were on the IMRL and they all retain the original frame numbers as built. I previously reported the WSOR 2052 wearing the frame #5602-B40. That would make it ex-WSOR 2008, exx-NREX 2008, exxx-IC 2008, exxxx-ICG 2008, originally an SD24B, UP 439B. (Lance Wales)

I've found a number of EMD units from mid-1940's instead of having the frame numbers stamped near the right front corner and left rear corner have them stamped several feet back from these corners. (Norm Metcalf)

Starting with the GP50 era they moved the number to the inside of the stepwell to the left of the second step. Still right front and left rear. (Russ Stodtz)

Quite a few of the more recent EMD units, instead of having the frame number stamped into the plating on the left side of the stepwell, have the number stamped into the vertical lip of the deck directly ahead in the stepwell, some frame numbers are stamped into the vertical lip of one of the treads. (Norm Metcalf)

"F" and "E" units were under the grab iron on the right rear between the drawbar and side of the unit. In essence at the "end" of the frame. (Russ Stodtz)

Above the drawbar on La Grange or Cleveland units is a cast or welded buffer. There will also be a number stamped there. The order number at that location is correct but the parts were not used in sequence. (Russ Stodtz)

GMD-London did not use frame number identifications until production was moved from La Grange. (Russ Stodtz)

GMD London, Ontario did not stamp frames with the C- number for units built for companies in Canada, like CP/CN etc. You can find the C-number on some newer power, but it is difficult to see without a flashlight, and they do not show a dash number with the C- number, so it is of little use in tracking the history. (Ken Ardinger)

These numbers are on a lot of later GMD built domestic locomotives, but as stated they are very hard to find and even harder to read especially if the locomotive is dirty, as most are. These numbers are in the I beam part of the underframe, not the outer frame where most frame numbers are stamped, and are almost directly above the front truck and almost in line with where the rail itself is. These numbers really tell the factory which frame is for which order, and they are not necessarily used in numerical or consecutive order. (Doug Cummings)

EMD Cleveland plant did not stamp frame numbers, that is why some early GP7's built there do not have frame numbers. (Ken Ardinger)

Nick Pitsch adds several points:

When you look at an EMD switcher, cowl or hood unit, look on the frame toward the right corner (no matter what side you are on) - on one side it will be near the "F", and exactly on the opposite corner of the other side. Sometimes you find it kind of inside the cutout where the steps connect to the frame. If you cannot read it on one side, try the other, and is always a good idea to check both if the opportunity avails itself.

Paint, junk, reflective striping and wreck repairs all can make reading the numbers difficult or impossible - it is best to note which numbers are "iffy" when recording them - sometimes a partial number can be ID'd thru process of elimination - only so many locomotives will match a "6?68-20", for example - and if it is a hood unit and the rest of the potential orders are E's or switchers, you have a likely match, and further research of rosters may serve to prove what you have located. When I record, I will underline a number fragment to remind me it check to see if it might be a "3" or "8", for example.

Be wary of Canadian produced locomotives from the late 1980's and back (if memory serves) - they didn't have 'frame numbers'. A clue is to look for "Dofasco" cast onto the truck - but these can be swapped rather easily with the US-made LFM/Rockwell trucks nowadays with US/Canadian locomotive intermixing - so this will serve as a clue only.

Don't bother to look on CF7's, as they used to be F units, as far as I know all evidence was obliterated when they were rebuilt. E and F units are on the rear right, low, even with the coupler. Haven't checked many of these personally, so it may vary.

In addition, order numbers weren't used before sometime around WWII - but it is worth the switchers that are still around from before then - they have been known to have "AO" numbers stamped in the 'wrong' corners... or all corners... or none at all. On some occasions, I have also found an order number stamped above the coupler on the buffer, but those numbers are, in my opinion, somewhat unreliable - but it can at least serve to ID what order it came from. Recording whatever you find, however, may serve to ID it, if not sooner, then later.

If you have access to the interior of the nose, EMD road units have the frame number stamped on the collision post in the nose. (Ken Ardinger)

With Alcos, the cab end bolster has the order stamped, but good luck in finding it. It is helpful to check the engine block plate. They do change blocks, but the number of the present block can help lead back to other units that may be from the same RR or dealer. (Ken Ardinger)

EMC and EMD engine block plates are useful, we can trace any EMC EMD built up to early 1959. (Ken Ardinger)

Engine blocks are frequently returned to EMD on an exchange basis, so could go back in anything. (Larry Russell)

Plymouths, built in the late 40's early 50's on, have the serial stamped in the frame similar to EMD's, but only on the rear of the units. (Ken Ardinger, Larry Russell)

GE and GM both stamp order numbers into other parts of the locomotives as well, again mainly so the shops/suppliers know which parts are for what order, these stampings appear above the couplers on both EMD and GE locomotives, and elsewhere. Again, these numbers only tell you what order the locomotive is part of as built. Over time with shop work, wreck repairs, etc., components get changed and if these components have numbers stamped in them this can lead to false assumptions so as time passes these numbers should be less and less relied upon. There are GM locomotives out there with two different frame numbers neither of which is correct for that locomotive, all a result of wreck repairs or shop work where a damaged section of the frame was replaced by a good piece. (Doug Cummings)

EMD Order Numbers

For an excellent summary and explanation of EMD's order number system, see Andre Kristopans' EMD Order Numbers web page.

EMD Serial Numbers

What is the significance of EMD's serial numbers from 1972 on?

The first two digits appear to be the year the order was placed, while the last two to four digits appear to be sort of a sequential order number. However, since 1975 the last four digits are usually 6xxx or 7xxx, so this can't be much of a sequence. (Richard Barnes)

The first two digits are the year the initial inquiry and bidding was done on the order. The actual order may have been placed then, or later. The order numbers are assigned at this time, even if the bid does not result in an order there is an order # assigned to it. The third digit indicates the type of order, i.e. if it is an export order that digit is an 8, components for locomotives to be built elsewhere get a 9. Domestic orders get a 6 or a 7. Marine power plants and other products get other digits. The last three digits are in sequence from 001 and up. They used to start these numbers at 001 at the start of every year but lately they have let them run awhile before going back to 001. (Doug Cummings)

The following comes from a source that worked in EMD's engineering department starting in 1968:

Doug is correct that an inquiry number is assigned as soon as the sales group requests it, so it may have a year one or two years ahead of when the order is delivered. The inquiry number becomes the order number once the details are finalized and the order placed. Within EMD, prefixes were used: I- meant Inquiry, A- meant advance, L- meant it had late delivery charges associated. Once an order was firm, the prefixes were dropped, except the L- which stayed with it thru delivery (but never appears on the builders plate) so people could focus on the importance of getting it delivered. 

The 6 or 7 Doug refers to denotes whether the unit is a regular production order versus an LRO (locomotive rebuild order). 6 meant all new, 7 meant some reused/rebuilt equipment, most often trucks and traction motor cores, sometimes crankcases or crankshafts. 9 was referred to as a "component set" order for an associate overseas. The component set could contain anything the associate wanted or needed but almost always an engine and alternator. (email dated May 12, 2007)

I've found a number of EMD units from mid-1940's instead of having the frame numbers stamped near the right front corner and left rear corner, have them stamped several feet back from these corners. (Norm Metcalf)

Quite a few of the more recent EMD units, instead of having the frame number stamped into the plating on the left side of the stepwell, have the number stamped into the vertical lip of the deck directly ahead in the stepwell, some frame numbers are stamped into the vertical lip of one of the treads. (Norm Metcalf)

What is the proper way to show the early builder numbers?

(For instance, Texas & Pacific NW2 #1000 had builder number E-742-1. Is there a dash between the letter and the first number? I have seen builder numbers published both with the dash and without.)

These are frame numbers, derived from the Order Number on EMD models. When stamped into the frame, they appeared in a variety of ways, but usually with a sideways "1" used for the dash. AFAIK, there is no "proper" way, since they were stamped in a variety of ways. The way you show it above is probably the most common and certainly understood. (Tom Lundeen)

GE Batch Numbers

Comparable to EMD order (or frame) numbers, GE used batch numbers, which were stamped into other parts of the locomotives, mainly so the shops/suppliers know which parts are for what batch (or order), These stampings appear above the couplers on both EMD and GE locomotives, and elsewhere. These numbers only tell you what order the locomotive is part of as built. Over time with shop work, wreck repairs, etc., components get changed and if these components have numbers stamped in them. Tthis can lead to false assumptions so as time passes these numbers should be less and less relied upon. There are locomotives out there with two different frame numbers neither of which is correct for that locomotive, all a result of wreck repairs or shop work where a damaged section of the frame was replaced by a good piece.

GE switchers had the serial numbers stamped in the carbody directly under where the builders plate was placed-but usually only on one side. (Ken Ardinger)

GE uses a "batch" number that is stamped into the buffer beam above the coupler. This batch number is repeated every few years, but is useful since very few batch numbers repeat over a different model. The number can be very hard to read because of paint of heavy impact on the front of the buffer plate. The newer units place this number on the top of the buffer plate. number typically should be in the range 1300-1 to 1499-99 for the B-B locomotives and 1800-1 to 1899-xxx for C-C locomotives. They have recently gone into the 1900's for C-C locomotives, which used to be reserved for electrics. Export units also have a number that corresponds to their weight plus the dash sequence. GE did not always deliver in sequence, so all numbers in a series must be checked. (Larry Russell)

GE road power has the batch number stamped above the front coupler, but they are not always in a numerical order for each unit built-so, it is good only if you have a list of how each unit was stamped when built. (Ken Ardinger)

As far as I know, I don't believe the serial number is stamped anywhere on the unit. The batch/purchase order number is still stamped into the front pilot, but, as noted before, these are not always sequential. With the first accident to the front pilot, this stamped number will likely disappear. Two notes: 1)  No batch number on the rear pilot. 2)  The new BNSF's are painted to the extent that the number becomes difficult to read through the green. (Steve Gerbracht)

The GE batch numbers on the buffer (until recently) and now on that 45 degree plate above and to the rear of the buffer (with some units having no batch number that I can find) are seldom numbered consecutively within the batches (which was almost always done consecutively by EMD's frame numbers). However, they're still worth recording, particularly with units missing their plates. They're one of the last resorts in identifying GE's. While the batch numbers have been recycled at least you can reduce the possibilities. Most recent four-axle power has batch numbers in the 1400s, most recent six-axle power has batch numbers in the 1800s and 1900s. (Norm Metcalf)

GE batch numbers would be the GE equivalent of EMD frame numbers. However, unlike with EMD, the dash numbers are in random, not numerical, order. The main purpose is to know which order the parts are for. But the numbers get re-used on a rotating basis so they do not have nearly the same significance. When they get to the end of the assigned group of numbers they go back and start at the beginning. All locomotive batch numbers start with a 1.1401, for instance, was likely applied to at least a half dozen, or more, different orders and will get reused every couple of years. (Doug Cummings)

GE Moser List

What is the GE Moser list?

The formal name, as shown on the list sheets themselves, is "Record Of Locomotives By Serial Number" and they were compiled by General Electric employees, for internal company use.

(Read more about the GE Moser List, including a link to the list itself)

Model Designation On Builder Plate

(Read more about locomotive model designations)

Is the model designation, i.e. GP9, SD40, SD70MAC, shown on the builder plate?

Whenever I give a locomotive model designation, I try to use what it says on the builder's plate. Different railroads call their locomotives different things - a CW44-9 on CSX is a C44-9W on UP and a DASH 9-44CW on BNSF (on the frame, anyway; I understand they use C44-9W elsewhere to save on column width) - but the plates will all say the same thing: DASH 9-44CW. In the fan press and on the net, the designation for Amtrak's 800's has been listed so many different ways that I've lost count, but if I have to give a designation, I figure I'm safe using what the builder's sticker says: DASH 8-40BP. (Evan Werkema)

For SP's early 8000-series, and LMX's 8500-series Dash 8 40Bs, these locomotives were some of the first "production" Dash 8's, and they also came out about the time GE flipped the nomenclature around. Santa Fe's pre-production Dash 8's (7400-7402) say B39-8 on the plate, and their production units (7410-7449) say Dash 8-40B. The SP and LMX units are supposed to be 39's, and I've sometimes seen them listed as B39-8E (E for enhanced) to differentiate them from the pre-production units, but what does the plate say? (Evan Werkema)

The plates on the LMX units designate the model as "B39-8". (Norm Metcalf)

GE used the Bxx-8/Cxx-8 up through the completion of production of the B/C39-8 series, including the so-called B/C39-8E. The change in settings to achieve nominal 4000 HP coincided with a marketing change which saw GE market their new locomotive line as the DASH 8 series. Their nomenclature changed at that time to Dash 8-40B/C and has remained so up through the Dash 9's. Numerous railroads have adapted their own classification nomenclature based upon their own preference and the available space in the computer. The placing of these classification titles on the locos seems to cause tremendous confusion among the fans, but it should be viewed as a owner classification, not a builder classification. One example of railroad-based nomenclature GE designations is that used by NS in its mechanical dept records, contracting Dash 9-40CW to D9-40CW. (Bob Graham)

Speaking of CSX and GE model numbers, we are still left with the conundrum of what to call CSX's 9000-class. GE at first said they were the first Dash-9s, and I think their plates do say "Dash 9-44CW". CSX has "CW44-9" stenciled under the road number. But my understanding is GE soon changed the paperwork to reflect the fact that these were really just 4400hp Dash 8s. I call them D8-44CWs when I label my slides. (Scott Chatfield)

I likewise believe the CSX 9000's are actually Dash 8-44CW, although CSX requested GE to call them Dash 9's from a promotional point of view, and it is true the plates call them Dash 9-44CW and the RR class shows as CW44-9. They lack several features consistent with the Dash 9 series: split cooling radiators; high adhesion "roller blade" trucks (prior to the GE steerable truck); and the longer Dash 9 underframe, which GE has confirmed as features of all Dash 9 locomotives. I call them D8-44CW when I'm lucky enough to get one, too. In addition, the 3 highest numbered CSX 7900's (I have seen and confirmed 7916) are marked on the plate and the cab RR classification as 4400 H.P. Dash 8-44CW/CW44-8. I recall one additional higher numbered CSX Dash 8 that is also rated at 4400 hp, but I can't recall the number. I did photograph it and can dig it out if necessary. (Bob Graham)

For SP's and D&RGW's SD40T-2s, I've almost always seen these listed as SD40T-2 and SD45T-2, but somewhere I recently read that the "T" was a railfan creation - EMD never used it in the model designation. After four years of living in SP country, I still haven't gotten close enough to a tunnel motor to read a plate - anyone know what they say? (Evan Werkema)

There are DRGW tunnel motors with plates saying "SD40T-2" though some of the early ones apparently only said "SD40-2". (Norm Metcalf)

The following comes from a source that worked in EMD's engineering department starting in 1968:

EMD's engineering documents list the model designations as SD45T-2 and SD40T-2; it is not a railfan creation. I have seen documents from other departments at EMD refer to them as SD45-T2 and SD40-T2 but that is not engineering's designation.

As information to how data was structured at EMD, which I don't see reference to on your site or others, each model when created was given a PLI number - PLI means Parts List Index. The PLI is the engineering document that is basically a parts list of all the major assemblies that are used to construct the locomotive. As far as I know, this practice dates to the beginning of EMC. The format of the PLI and numbers changed over the years - in ways I don't understand since I never asked. For example, an F3A had 28 different PLI numbers over it's production run starting at 2700 and ending at 2837 and the F3B had a similar 28 starting at 2701 and ending at 2838. The F7A only had two PLI numbers as did the FP7A and F7B so either the locomotive configuration stabilized or a new PLI was being assigned for each order at the time of the F3, which I think is likely. Four digit PLI numbers were used until the Dash-2 series when the numbers went to 5 digits starting at 92000 which is the PLI number for the SD45T-2. PLI numbers were assigned sequentially and by the time I retired were over 92145 meaning 146 different models of locomotives were designed or at least contemplated over that 35 years. It made no difference whether it was a domestic or export model, it got the next number in sequence. For reference, there are a few interesting models that were assigned PLI numbers but never built, such as 92061, the GP50T which was to be a tunnel version of the GP50. During this period, each model only had one PLI number.

The PLI did not represent a complete locomotive, however. Recognizing that different customers would want different options, certain areas of the PLI list a part name but for part number they show "SEE SPEC". In this case, the customer order number generated a listing of the options that applied to that order. On Dash-2's, the most significant was dynamic brake, for example. At EMD, options were called EDL's, Exception Drawing List. Each option got an EDL number assigned; the EDL defined the part numbers making up the option and deducted any standard material if required. EDL's were 5 digit numbers when I started and retired but had gone from about 20000 to 60000 indicating 40,000 options were developed over those 35 years. Tying this all together was a document called a BLS, Basic Lomotive Specification. The BLS resembled a customer order as it was a reference to the PLI plus a listing of all the EDL numbers that created a complete locomotive. The BLS was used as a starting point to establish a basic locomotive for features and pricing.

Within the Engineering department at EMD there is a vault where the engineering records are stored. The vault housed all the drawings and all the order paperwork that engineering used or generated. For each order actually built, there was a folder, filed by order number, that contained the specifications (PLI number reference plus a list of EDL's) and correspondence between Sales and Engineering defining the order, along with the final specification release that lists all the EDL's the order was built with, and including the table correlating road number and order/unit number as delivered. The drawing reference for the painting and styling is also contained in this record as well as the specified weight. As of my retirement date, the vault was still complete back to the earliest days of EMC, but it would not surprise me to find it purged of early information that may not have commercial use today, expecially with new owners who have no sense of history. I used the info in there throughout my career on numerous occasions and was always fascinated by it. I hope if EMD purges it, someone like the Illinois Railway Museum can retain it. (email dated May 12, 2007)


What date is shown on the builder plate?

There are several dates important in the history of any one particular locomotive:

For your rosters, always use the date as shown on the builder plate. But keep in mind that sometimes, the builder date and the delivery date can be different due to problems in financing and actual delivery schedules. One twist to this is that roster keepers have now documented that many modern GE locomotives have had their builder plates (actually decals) changed, with different build dates shown, i.e. Sep 1993 changed to Aug 1993. The reason for these changes are not known, but likely it has to do with bringing the build date in line with the date on warranty documents.

All builder's dates in any roster are suspect. Actual plate checks were rare then, and are even more rare now. That is why Don Dover 30 years ago, began using the exclamation mark (!) to show an actual plate check.

There is another date on every locomotive -- the blue card date, located on the FRA blue card in the locomotive's cab. The dates I call in-service dates are the cab blue card dates, meaning the date that the air brakes become active for FRA inspection purposes. But actual circumstances have proven that the blue card date can often be wrong, or even missing completely.

While the delivery dates and on-property dates are interesting (and usually different by as much as a week), it is the build date on the builder's plate, and the blue card date (also known as the air brake date) from the unit's cab that should be of most importance.

For roster purposes, the build date should reflect the date as taken from the builder's plate (or decal), and most of the time, this is divined from multiple sources because plate checks are difficult to get for complete orders. The word processing software I use (Microsoft Word) allows multiple columns, so on new units, I could add delivery dates, or in-service dates. But this only confuses almost everybody, so it usually boils down to a generic build date, which could be any of the above dates, i.e., builder's plate date, delivery date, in-service date. Delivery dates are usually only available from the builders. In-service dates are usually only available from the railroads, and usually only in the first year or so following delivery.

Of course, this always brings to mind the day-dates that EMD put on its units in the 1940s and early 1950s. Was that day the same as today's blue card date? Or was it some other date? Don Dover (the editor of Extra 2200 South from 1968 to 1990) and I agreed many, many years ago that it was most likely the equivalent to today's blue card date, meaning the day that the air brakes became active, and so many tons of steel and other materials officially became a locomotive. But that would mean that the unit sat waiting for its builder's plate to be completed before delivery could take place.

Always check the builder's plate. When possible, check the frame number. Also, if possible, get the blue card date and previous unit number. The previous number may not be correct (by our records), but it is a start.

This is an on-going discussion for roster keepers, and explains why plate checks are so important to locomotive fans. But to the builders and the railroads, it is the in-service date (for warranty purposes) and the blue card date (for FRA safety) that are most important.

In 1972-1978, while I was working at UP, and building the UP roster that was in X2200 in 1979 and 1980, I was going around the Salt Lake City shop and service track every morning doing plate checks (and frame checks when possible). I got a lot of questions, strange looks, and some pretty unusual comments from my fellow workers.

I have all the delivery dates for UP units back to 1986, but I have always used Builder Date out of tradition. Coming up with those magic builder's dates is getting harder every year. From my sources at UP, I know that the build date is not important, except by year. And the builders themselves depend on delivery dates (as do the railroads) for warranty reasons. Maybe the date on the builder's plate is only a minor detail to both the builders and the railroads.

In reality, the two dates are usually within the same 30 day time period, with maybe a week after, making the delivery date possibly one month later than the build date. With the railroads being so sensitive today to trespassers, getting actual builder's dates from the builder's plate on each unit is getting harder. I originally only had actual builder dates on about a third of the fleet of the new UP SD70Ms.

Builder Plates

As far as I know, including the ten years that I worked for UP repairing locomotives, the company itself, and its management in Omaha never had any interest in builder plates, and did not care if they were removed or not, whether in service or when being sold for scrap. The company kept records concerning build dates almost solely for warranty purposes, and there were numerous contemporary sources available whenever a concern developed. There is no reason to need the information after a locomotive has been sold or scrapped. Many people have seen what they describe as piles of builder plates in locations throughout railroad company property, including locomotive repair shops. As far as I know, these were retained merely as mementos for managers and retiring employees. and were sometimes

The build date of a locomotive, and its serial number, as shown on builder plates, has only been used by railfans and locomotive historians. Builder plates are used by these two communities merely as mementos, with a brisk market developing for particular plates of particular design, or from particular locomotives.

My own research concerning the accuracy of so-called builder dates, in which I was able to compare company records with actual builder plate information for close to a hundred locomotives, shows that the company records in-fact show delivery dates, or in-service dates, not builder dates as shown on builder plates. Delivery dates and in-service dates are very important due to warranty issues, and to prove collateral for loans and trusts, but actual build dates are of little concern.

Additional Comments About Dates

Be *very* suspicious of dates of loco retirement, scrapping, sale, etc., that are the first or last day of the month, year, tax year, etc. These are often the date when these events were officially authorised by higher management, not the date when the event actually happened. (Brian Rumary)

One extreme example of builder date vs. delivery date: EMD built a group of MP15AC and GP38-2 units for NdeM in 1983 but due to financing problems the units sat in EMD's LaGrange backlot for a couple of years. (John Benson)

Builders plate dates are one thing you can obtain from a locomotive without having access to the railroad computer - and unless the plates/decals have been stolen or obliterated should remain there for the life of the locomotive. On the other hand, delivery date information is only available from the railroad or the builder and only if you have access to that information. (Doug Cummings)

The delivery date is a "real time" date that is purged from computer systems pretty quick, say about 60 days is the current policy. On the other hand the air or FRA date is supposed to be recorded and available on a long term basis. 30 years ago the builder's plates were a great resource. Times have changed. (Russ Stodtz)

Ian Platt adds some thoughts about four dates used for a new locomotive:

Curt Jans adds:

When I was a co-op with Conrail many years ago, they had a build-date in their computer system, but it was actually an in-service date. In the case of many older units from the predecessor railroads, the data was only as good as the data recorded by another railroad 30 years earlier.

On BNSF, each locomotive has a "date new" line in the computer system. The date new line reflects the first inspection/air test date entered on the blue card when the locomotive was delivered. As far as BNSF is concerned, this is the gospel, with respect to maintenance intervals.

On CSXT, the publications department produced a "roster book" every eight years or so. It had a line-by-line entry for each locomotive including original road number, serial number and build date. Once again, this data was only as good as the data created/collected by the predecessor railroad. In the case of the L&N, the build date was strictly a year (1965, 1970, etc). I assume that L&N officials never noted the month, and therefore, CSX used what data they had.

Bob Lehmuth comments:

Going back to post WW2, Vulcan Iron Works (apparently) cast a large number of builder plates - probably anticipating a postwar "boom." Problem is, it never happened, so the last locomotive built by Vulcan - in 1955 - carried a 1948 "build" date.

This involves approximately 100+ locomotives. Taking data from builders as gospel is [can be] a MAJOR mistake. This opinion comes from my compiling 50+ builders' records and finding most builders are/were pretty sloppy - to be kind.

What exactly is a builder's date anyway?  Is it the air brake test date?  A contractual deadline?  Roll-out date?  Shipped off the property date?  Paint dry date?  Builder plate stamped date?  Paperwork done date?   What do the builders use?  Is it consistent? In my B&O roster, I just use month/year for builders date. Information from rebuilders (i.e. M-K/Boise Locomotive) often comes as month/day/year so I use it. I assume it is the nothing-left-to-be-done date. (James Mischke)

It's been previously publicized (though I don't recall the unit) regarding a difference of about three years between the builders date and the delivery date. Also, what about demonstrators? Some of these have spent considerable time cruising around North America before being sold. (Norm Metcalf)

The delivery date is not recorded on the FRA blue card. BNSF does not even use the concept of a "Built Date". The nomenclature is "Date New". This would be the date of the first FRA inspection and the first FRA mandated test of the air brake system. This is the only meaningful date since the future maintenance schedule for the life of the unit pivots around those two dates. I record both but the "Date New" is the important one. (Russ Stodtz)

Additional comments from Russ Stodtz:

General Electric faxes the BNSF FRA/Air dates as the units are shipped. As far is BNSF is concerned that is the Builder Date if, like in the UMLER, there is a box to put such a thing. That date becomes the "Date New" in BNSF records. I think the KCS also uses that same convention.

One more important thing about that date. While it really does not matter much considering the numbers involved but the "Date New" does trigger the delivery grace period. CR/CSXT have 7 days from that date before the unit goes on the HP/Hours account.

There were a few units that were not painted at Erie that either did not have or lost their blue cards. There were also a few units that showed up at Cicero without blue cards. In both cases the "Date New" would be the date the blue card was signed, whether at Mid America or by the diesel shop Foreman at Clyde or wherever.

In the case of the BNSF C44s, many are then routed to Thunderport Industries in Corry for paint. The actual in-service date is often two to three weeks off from the actual build date. Likewise, when the first batch of SD90MAC's was built for UP, many spent months at VMV for tune-up purposes. The build dates stamped into the plates of these 90's reflect their build month and not when they actually entered service on UP. (Curt Jans)

Three examples from Andre Kristopans:

On at least 3 batches of GE's for ATSF in 1980's that were built before financing was finalized, and so sat at Erie for several months before being shipped. In addition, a couple of batches of EMD's for Mexican National Railways sat for nearly a year for similar financial problems.

A second problem can occur with locos built for stock and sold later. Most noticeable case I can think of is early EMC SC's and SW's, which in some cases (like EJ&E SW's 200-208) were finally sold over a year after construction! Problem here is that plates were frequently blank originally and were stamped with the current date when they were being prepared for sale.

Finally, there is the matter of former demos. Extreme case: The E6's built for display at 1939-40 World's Fair and sold to SAL. An A and a B were built early-1939, the A was sold to SAL 01/40, and a new A was produced for 1940 fair season. Circa 10/40 this 2nd A plus the original B (now a year and a half old) also went to SAL. To just show "delivery dates" in this case would be highly misleading.

Until a few years ago, "EMD" meant built, completed, painted, everything at LaGrange (except for units built at Cleveland, but more on that later). Now, you have EMD building locos at London, Super Steel, CP Angus Shop, Sahagun, Juniata, and I am sure more to come, then shipping them to VMV or wherever for painting. This stuff will NOT be possible to reconstruct from existing records in five years unless it is noted now. Remember, computer records are purged a lot more frequently (and irretrievably) than earlier ledger books. If you do not think this will be a problem in the future, unless it is addressed while current, look at the EMD Cleveland situation. I am sure that in the late 1940's and early 1950's the question of what was built where could have been answered relatively easily, but nobody at the time figured it to be important, and by the time people started asking questions, it was too late to answer them. Is it really possible that ALL SW's and GP's in the affected time span came from Cleveland? As far as I have been led to believe, the Cleveland plant was quite small compared to LaGrange. (Andre Kristopans)

Several points from another source:

I think builders plates became relevant/important because they are a physical entity on the side of the locomotive for life (one hopes) that contains the Build Date information ... no checking computers, no getting into the cab.

I keep track of the actual "delivery date", i.e. the day the unit is interchanged to the owner railroad (in other words, the date the unit hits the rails of its owner for the first time). Builder plate may say 11/00, delivery date in the computer may say 11/15/00, but if the unit gets trapped by CSX or some other railroad for a month the actual delivery date (when it gets to UP say) may not be until 12/31/00. By the way, this might differ from in service date as well, which could be the same date as the unit was interchange received or up to a week later (could depend on when the builder's rep can get there to sign off with the appropriate RR personnel). Could also be in a completely different location - what if a unit is received at Chicago, but won't be accepted until it reaches North Platte?

Delivery date via air brake test is more of a "sign-off" date it seems or an acceptance date than an actual delivery date ... I think "delivery date" in this usage is inappropriate.

With contract shops doing a lot of the painting work for both builders Build Date is pretty irrelevant for many units. Take UP 4020 ... the unit was built at London in 4/00 but by the time VMV gets around to painting and releasing it it is 9/27/00 ... and the unit was interchanged (delivered) to the UP on 9/30/00 -- quite a time span. How do you keep track of that one? What's the air date on it?

There are several orders since the '70s that were built but not delivered for a year or more due to these problems.

From a source at GE:

For my notes, I keep:

1) builders plate info (builders number, build date)
2) frame/batch number
3) delivery date (from GE to connecting railroad)

I have not dug for the acceptance date of each individual locomotive-typically the units are inspected and "signed off" in the builder's final test facility, but I have never pursued the paperwork to document this.

Often, especially these days, they are off the property as quickly as possible following sign-off. However, this may not always be the case if units are then stored due to lease/financing agreements. For example, the order of 15 Ferrosur AC4400s were built/painted/tested/signed off as complete in March, but were then put in cold storage behind the plant until final lease/financial agreements could be signed off.

Overall, the builders date on the plate is still generally accurate to construction date (+/- a week or so).

As explained to me by a GE contact a few years ago the "pre-delivery" of locos prior to what the date of the sticker says happens from time to time and is because the stickers are pre-printed, sometimes even in advance of construction, but basically the dates on the stickers are the expected build date, and sometimes GE gets ahead of themselves in terms of turning locomotives out. (Sean Graham-White, March 26, 2008)

Dates Used By UP

The date UP uses is the interchange date, or the on-property date (the date the unit first enters UP property via interchange, as recorded in their TCS traffic management computer). This is the date that is entered as the "Additions" date for a new unit on the monthly changes sheet that is broadcast to all TCS users.

The dates that UP keeps track of is the year built, and the blue card date, for FRA reasons. The builders most likely only care about delivery dates (date turned over to the owning railroad), for warranty purposes. In the case of new GEs, UP, GE and Conrail may have an agreement that allows GE to make official delivery to CR, and CR operates the unit as an official UP unit in pool service.

On UP, a paper copy of the blue card is mailed from the builder to the railroad at Omaha, and the date on that copy is used to enter an initial date into the automated blue card system (part of the Locomotive Management System) to drive the next inspection due date, then the paper copy is filed for future reference.

Trade-In Units

Between 1953 and 1962, UP received 20 passenger units--E9s 900-914, E8 925, and E8Bs 922B-925B--from EMD that were built using direct unit-for-unit trade-ins from UP. Railroad and EMD records both show the trade-in units as being rebuilt to the respective E-units. This rebuilding of an old unit to a new unit is unlikely because the mechanical layout of the early E-units and the "rebuilt" E8s and E9s, were completely different, with almost nothing reusable except truck frames and some minor parts. The rebuilding concept stemmed from a tax advantage that allowed UP, and several other railroads, to pay a reduced tax on the units, since they were "rebuilt" rather than "new." This tax loophole was closed after many industries nationwide began receiving large amounts of "rebuilt" equipment, rather than new equipment.

In the case of railroad locomotives, each unit-for-unit direct trade-in came under EMD's "trade-in credit" policy, which was figured strictly on paper. Due to the short length of time between the trade-in and when the new unit was delivered, and the mechanical differences, it is unlikely that any one part from the trade-in unit was actually used on the intended new unit. It was EMD's practice during E8 and early E9 production to assign the trade-in units' builder's number to the new unit, and to show the trade-in unit's original build date on the builder's plate, even if the new unit was built using as little as five percent of the parts from the older unit, thus UP 922B-925B, and UP 925, were built as new units but were assigned builder's numbers and builder's dates from the original trade-in units. The roster shows the actual build date rather than the build date from the builder's plate.

It is common to see the fiction in print that B&O (and ATSF and GM&O) had EMD rebuild their early E-units into E8m's. Truth is that so few parts were used that this event was a trade-in. On B&O, only the main generator, horns, fuel pump, and several other parts were reused on the new unit. Trucks, frames, and prime movers were scrapped. EMD and B&O maintained this rebuild fiction for tax and accounting reasons: the "rebuild" could be expensed off quicker than a new unit. Corporate management everywhere will bend the facts to save cash on taxes. Little EMD documentation survives, they systematically destroyed almost everything. EMD had been charged in federal criminal court (case open 1961-1966) for unfair trade practices, by threatening railroads to withhold GM automotive traffic if the railroads did not buy EMD locomotives. This was successfully dodged by stonewalling the discovery phase. No sense in letting you or me years later to find the evidence, so successfully withheld. (James Mischke)

EMD Anti-Trust Case

Andre Kristopans asked in April 2003:

How did the GM anti-trust case result in more EMD data coming into view? Seems like the single most important source of EMD serials were the Product Reference books, and given that they existed since at least 1942, as I understand, I am sure a few of them "escaped" as far back as the 40's.

Jerry Pinkepank's giant contribution was the "Diesel Spotter's Guide", the first attempt at systematically defining locomotive model designations. Until then, it was almost a case of everybody "having their own idea" of what a GP9 looked like vs a GP7 or GP18.

Larry Russell responded:

Jerry worked on the anti-trust case and managed to break free the first copy of the Product reference book with serials as well as Alco. The anti-trust case was the key to obtaining the information, until then, we were compiling from sightings only.

James Mischke responded:

This was in 1961, and proceedings dragged out until 1965.

The essence of this criminal case was that EMD leveraged its massive high tariff freight traffic to armtwist railroads into buying EMD locomotives. The Justice Department held grand jury proceedings and criminal indictments resulted.

One charge was that EMD threatened to close the Bridgeport, Conn. refrigerator (Frigidaire) plant if New Haven did not buy FL9's for its passenger service instead of competing products (i.e. F-M Speed Merchant). Close to bankruptcy and unable to risk that loss of traffic, New Haven buckled and bought FL-9's.

I have been very interested in this case and looked up the district court case transcript at the National Archives branch in Chicago. There are about seven solid feet of paper, which I went through. Mostly lawyers arguing the law in open court. Not the facts.

The specifics are in the preceeding grand jury testimony, which are forever off limits to anyone other than US attorneys, as I was told when I asked to see them.

This was right in the middle of the Monica Lewensky affair, when privileged grand jury testimony was on the network evening news every night. These news leaks were completely illegal although no one was ever punished. The archivist was unimpressed by this recent legal precedence, or the lack of punishment, and I still could not see them. There is juicy stuff in there, mostly probably hearsay, permissible before grand juries. Not available to us railroad historians..

Bottom line is that General Motors stonewalled until the government gave up. The discovery process was thwarted by GM at every turn (we don't know where that manager is anymore, he retired; we don't have those records anymore; which pages do you want again? Whaddaya mean you want the whole file copied, that's unreasonable ......). It never came to trial.

No wonder EMD has been completely uncooperative with potential EMD historians (Marre, Hirsimaki, Mischke), there are skeletons in those closets.

The only good things that will ever come out of that fiasco, are the late nights law intern Jerry Pinkepank spent in front of the copier (then called a thermofax) machine at the Justice Department. That research was the basis of the Diesel Spotter's Guide.